Research On Image Colorization And Itsapplications | Posted on:2012-03-13 | Degree:Doctor | Type:Dissertation | Country:China | Candidate:Z Jin | Full Text:PDF | GTID:1118330362953673 | Subject:Computer application technology | Abstract/Summary: | PDF Full Text Request | Colorization is a computer-assisted technique of adding color to images and video. Existing work on colorization can be roughly divided into scribble-based colorization and example- based colorization. It has been widely used in industrial fields such as movie making, digital entertainment and cartoon making. This paper introduces three effective image colorization algorithms based on the study and analysis of image colors and its relationship with user scribbles, image structures and color emotion.First, a novel intrinsic-distance-based recolorization technique using scribbles propagation is presented. It requires less manual effort by taking full advantage of the original image. The scribbles are well relocated by propagating the user adjustments. A variety of results indicate the proposed method works effectively and avoids a significant user effort. As well as makes the colorization result is no longer depending on initial position of color strokes. An efficient framework for high resolution image colorization using this strategy is proposed to avoid the loss of detail in down-sampling.Color and structure are the two most important features in image. The orientation of sketch in comic images contains significant information for vision. A scribble based method to colorize comic images is proposed. It is guided with the orientation field which is generated from the primal sketch. The pattern-based features are measured by using the statistical features. The results indicate that the directional sketch model provides a straightforward way to present fine details in comic images, and brings significant time savings by propagating colors over various complex texture regions.Color emotion proves to be useful in colorization by psychophysical experiment. The method based on the color-science-based formula is developed to enhance the desired colorizing result by modifying the colors of user scribbles. The results demonstrate that our method is capable of automatically enhancing the color emotion with aesthetically pleasing in terms of human visual perception. | Keywords/Search Tags: | user scribbles, edit propagation, primal sketch, texture analysis, color emotion, color harmonization, image colorization | PDF Full Text Request | Related items |
| |
|